Professional Insight

Internet of Things: A New Flood of Data Awaits

In today’s interconnected world, it seems like anyone can keep track of anything from anywhere. That’s an exaggeration, of course, but the ability to remotely monitor via the internet promises to change how insurance policies are priced and underwritten and how claims are handled.

That’s the promise. And while it seems clear that the Internet of Things (IoT) will affect insurance, there are some questions regarding how great the impact will be.

A pair of speakers examined the potential implications at the 2017 CAS Spring Meeting in Toronto in May. One looked at homeowners insurance, the other at workers’ compensation.

There are about 2.3 billion items — e.h., cameras, refrigerators, blenders — that are plugged into the internet, and estimates say there could be 13.5 billion by 2020, said George Hosfield, a senior director at LexisNexis Risk Solutions.

He laid out the insurance potential of IoT with the story of a colleague whose teen child flushed a wad of paper towels down a second floor toilet that had a faulty shutoff assembly. The ensuing flood poured into the kitchen, causing a multi-thousand-dollar claim.

“It was a major thing that could have been easily prevented with technology today,” he said. An IoT water moisture sensor on the bathroom floor would have detected the water earlier and pinged a smartphone, preventing (or at least mitigating) the claim.

There are other examples, like a smoke alarm that pings a smartphone when it goes off.

Insurers have several ways to raise interest in such devices, Hosfield said.

  1. They can co-market with the device provider or give the device to their customers as a promotion, where the law allows.
  2. They can offer an early-adopters discount, anticipating that the device will lower claims or that early adopters of safety technology will tend to be safer risks overall.
  3. They can pay for the mitigating device themselves, expecting a net cost savings.
  4. They can rate based on confirmation that the device is present and active.
  5. Finally, they can employ an actuarial approach to the live data sent by the device, using it to rate by identifying patterns of behavior that correlate to the insured’s risk profile.

Right now most carriers that are engaged with smart homes are between steps 2 and 3, Hosfield said.

In the long run, how much of a discount is possible? It depends, of course, on what consumers choose to monitor.

Consumers prefer anti-theft devices, he noted, but those make up a relatively small portion of homeowners insurance non-cat loss costs. Most losses come from fires and non-weather water claims, like that overflowing toilet, but people aren’t buying sensors that monitor those.

The promise today is that an IoT device could alert managers when an employee gets hurt or — even better — determine when a hazard looms.


Another challenge for insurers will be sifting through all the data those billions of devices are collecting. Not only is there a lot of data, there is no guarantee that the data feed from one brand coffeemaker, for example, will look like the feed from another.

It may make sense, Hosfield said, for insurers to partner with a data platform that translates the disparate information into a single feed.

While home monitoring is just starting to establish itself, worker monitoring via IoT promises to be the next phase of a generations-long trend of technology making safer workplaces, said Peter Rousmaniere, a journalist who has written about workers’ compensation insurance for decades.

In manufacturing, for example, the number of serious injuries declined 63 percent from 1993 to 2012, a falloff only partly explained by the decrease (33 percent) in manufacturing jobs over the same time period.

The promise today is that an IoT device could alert managers when an employee gets hurt or — even better — determine when a hazard looms.

The improvements not only reduce injuries, he said, they increase productivity — an added incentive to employers.

“Behind every work injury is a productivity advance trying to get out,” Rousmaniere said.

He focused on hospital injuries — specifically, those caused by lifting patients — as an example of how new technology could work. The device could intervene at four stages:

  • Prediction — telling in advance that the patient will be difficult to lift.
  • Real-time monitoring — determining if the hospital aide is moving the patient correctly.
  • Assistive — indicating if more than one person is needed to lift the patient.
  • Robotic — eventually, replacing the worker with a robot or other device.

An insurer would want to have ways to analyze and monitor each of these, Rousmaniere said.

Hosfield, the homeowners expert, listed these steps for insurers and actuaries to learn more:

  • Get engaged, perhaps by getting an IoT device, to see how it operates and what data it gathers.
  • Work with data partners to determine how best to mine data.
  • Be realistic about what the IoT can deliver. “It isn’t going to be a panacea, but it will be useful.”